Skip to main content
Back to Blog
Compliance Tips

AI Act prohibited practices from 2 December 2026 (Article 5(1)(ba) and (bb))

18 November 20268 min readObligoBoard Team
AI Act prohibited practices from 2 December 2026 (Article 5(1)(ba) and (bb))

The AI Act prohibited practices that apply from 2 December 2026 are two new points in Article 5 of Regulation (EU) 2024/1689, points (ba) and (bb), inserted by Regulation (EU) 2026/1744. They concern AI systems that generate or manipulate non-consensual intimate material of an identifiable person, and AI systems that generate or manipulate child sexual abuse material.

This article describes the text and the evidence worth keeping. It does not assess your organisation's situation. The consolidated text of Article 5 can also be read on the Commission's AI Act Service Desk.

The two new practices, Article 5(1)(ba) and (bb)

Article 1(7)(a) of Regulation (EU) 2026/1744 inserts the two points into the first subparagraph of Article 5(1). Article 113, third paragraph, point (a), as replaced by Article 1(40)(a), provides that Chapters I and II apply from 2 February 2025, "with the exception of Article 5(1), first subparagraph, points (ba) and (bb), and Article 5(1a) and (1b) which shall apply from 2 December 2026".

Point (ba) covers intimate parts and sexually explicit activities

Point (ba) prohibits "the placing on the market, the putting into service or the use of an AI system that generates or manipulates realistic images, videos, audio or similar material of an identifiable natural person's intimate parts, or of an identifiable natural person engaged in sexually explicit activities, without that person's freely-given, specific, informed, unambiguous and explicit consent for that generation or manipulation" (Article 5(1), first subparagraph, point (ba)).

Recital 12 explains that this "'realism' refers to the depiction of the person's face, voice or their body in a credible real-life manner, regardless of the realism of the context of that depiction or whether it fully corresponds to the actual voice or appearance of the depicted person". It "excludes cartoonish or physically impossible depictions of a person's body". The recital adds that the prohibition does not cover generative AI applications where intimate parts are not exposed or, if exposed, where this is subject to the "freely given, specific, informed, unambiguous and explicit consent of the depicted person", and gives as examples "try-on applications and medical applications, such as medical anatomical simulations and mammograms".

Paragraph 1b, for point (ba) only, sets a limit on what counts as manipulation: "an AI system that manipulates material in a way that does not increase the exposure of any depicted intimate parts or alter the nature of any depicted sexually explicit activities shall not constitute manipulation" (Article 5(1b)).

A note on language versions. The English text of point (ba) requires consent that is "freely-given, specific, informed, unambiguous and explicit". The Italian text of the same point reads «liberamente prestato, specifico, informato e inequivocabile», with no word for "explicit", whilst recital 12 includes it in both languages. All language versions are equally authentic. For how to read this requirement, consult your adviser.

Point (bb) refers to Directive 2011/93/EU and a 'without right' defence

Point (bb) prohibits "the placing on the market, the putting into service or the use of an AI system that generates or manipulates material or performance within the meaning of Article 2, points (c) and (e), of Directive 2011/93/EU, except where a 'without right' defence applies under national law" (Article 5(1), first subparagraph, point (bb)). The definitions it refers to are in Directive 2011/93/EU.

Recital 13 states that the prohibition should not prevent the placing on the market, putting into service or use of an AI system "where a 'without right' defence applies under national law, as referred to in Article 5(1) of Directive 2011/93/EU". Among its examples is "the legitimate use of the AI system in the context of red-teaming and evaluation activities for the purpose of assessing the system's compliance with the prohibition". Whether such a defence applies depends on the applicable national law, which this article does not examine.

Providers and deployers, paragraph 1a

Point (a), which recital 12 presents as concerning providers, provides that placing on the market or putting into service "is only prohibited where" one of two conditions is met.

  • Point (i). "that generation or manipulation is the intended purpose of the AI system".
  • Point (ii). "the system's design, training, architecture, capabilities or user-facing functionalities make that generation or manipulation a reasonably foreseeable and reproducible outcome, without requiring significant technical modification, and the system does not have reasonable and adequate technical safety measures and other safeguards to reliably prevent that generation or manipulation, taking into account reasonably foreseeable misuse, and to correct observed or reported misuse".

Recital 12 lists measures and safeguards that "could include data cleaning, refusal training, safe prompt design and output controls, runtime prompt guardrails, content classification and filtering mechanisms, usage restrictions, abuse detection mechanisms, and notice and action mechanisms". For systems intended to generate or manipulate intimate material, it refers to "appropriate means for the distribution of the AI system aimed at enabling the reliable collection and demonstration of consent of the depicted person", in compliance with Regulation (EU) 2016/679.

For deployers, point (b) provides that the use of such a system "is only prohibited where the deployer uses the system for the purpose of generating or manipulating such material or performance". Recital 12 adds that the prohibition on use does not cover "accidental generation or manipulation of such content".

Also on 2 December 2026, Article 111(4) and Article 50(2)

Article 1(39)(b) of Regulation (EU) 2026/1744 adds a paragraph 4 to Article 111: "Providers of AI systems, including general-purpose AI systems, generating synthetic audio, image, video or text content, that have been placed on the market before 2 August 2026 shall take the necessary steps in order to comply with Article 50(2) by 2 December 2026." This is the marking obligation of Article 50(2), with the same date. It is separate from the Article 5 prohibitions.

2 December 2026 at a glance

ProvisionWhat it providesWho it concernsFrom
Art. 5(1)(ba)Prohibition on realistic material of an identifiable natural person's intimate parts, or of that person engaged in sexually explicit activities, without the consent the text requiresWhoever places on the market, puts into service or uses the system, on the terms of para. 1a2 December 2026
Art. 5(1)(bb)Prohibition on material or performance within Art. 2(c) and (e) of Directive 2011/93/EU, except where a 'without right' defence applies under national lawWhoever places on the market, puts into service or uses the system, on the terms of para. 1a2 December 2026
Art. 5(1a) and (1b)Conditions for providers (point (a)(i) and (ii)) and for deployers (point (b)); para. 1b states, for point (ba) only, what does not constitute manipulationProviders and deployers2 December 2026
Art. 111(4) (Art. 50(2))Necessary steps for generative systems placed on the market before 2 August 2026Providers of AI systems generating synthetic contentBy 2 December 2026

What to document

Points (ba) and (bb) and paragraphs 1a and 1b list no documents to keep. Recital 12 does speak of preventive measures and of the reliable collection of consent. Each item below is a process choice, tied to the words of the provisions.

  • Role and intended purpose, per system. Whether the organisation is a provider or a deployer, and the purpose for which the system was designed or is used. This relates to paragraph 1a, point (a)(i), and to use "for the purpose of" under point (b).
  • The technical measures and their date, if you provide a system that generates images, video, audio or similar material. Which measures are in place to reliably prevent that generation or manipulation and to correct observed or reported misuse, and since when. This relates to paragraph 1a, point (a)(ii).
  • Usage rules and their date, if you are a deployer. The internal instructions on what the system may be used to generate or manipulate, and the date they were issued and communicated. This relates to paragraph 1a, point (b).
  • Consent records, if the system is intended to generate or manipulate intimate material. These relate to the depicted person's consent required by point (ba) and to the reliable collection of consent mentioned in recital 12. The text does not say that this consent is the same as consent under Regulation (EU) 2016/679, and the records are themselves personal data.
  • The date of the review. The day the system was reviewed against the new points, by whom, and the date of the next review.

Where this sits in ObligoBoard

In ObligoBoard, the "EU AI Act intake" questionnaire adds the "Prohibited AI Practices Screening" obligation for providers, deployers, importers and distributors, due on 2 December 2026. The obligation quotes the new provisions word for word from the Official Journal and is tracked with its article reference and due date. Its evidence field asks for the systems reviewed, the outcome and reasoning per practice, the reviewer and the date.

Regulation (EU) 2026/1744 also replaced Article 4, covered in AI literacy under Article 4 of the EU AI Act.

ObligoBoard isn't open yet. See what ObligoBoard is

General information, not legal advice or an attestation; drafted with AI assistance. The official text published in the Official Journal of the European Union prevails. For any question of interpretation, consult your adviser.

Related Posts